In the story I read about the Space x rocket explosion, it said the satellite cost 150M.
How much was the rest of it worth? Rocket, fuel, launch pad, man hours, just wondering.
Pat said..."We all make mistakes, but even more so when the people are not qualified to be doing what they are."
I say....Or deliberately willing to IGNORE 7+ decades of best practices in the testing community...all well and thoroughly documented
Same with a lot of this stuff, the worst case of hubris probably being the second stage of Falcon 1 going unstable. It is definitely a hard problem to stabilize a gimbal-controlled rocket stage throughout an entire fuel burn, the inertia changes by a factor of 10, the mass/acceleration by a factor of 5 or so, and thus the slosh effects vary widely over the mission. But, one thing I do know, is that running 100,000 simulations, tweaking the gains by eye until it looks about right, is not a robust solution. That's why the "dinosaurs" cost a lot and take a long time, everbody did that already and found out the hard way.
Thanks for a frank report, man. I'd love to see that vid with the sound syncronized but the evidence as presented is very impressive.
I simply cannot imagine a static test outside of a test environment/procedure either, ie payload on top is dumb. If I owned the payload I would've bitched big time knowing that plan! So dumb.
As dumb as the Rutan guy unlocking the wing in highspeed flight;
Or ground level pressurizing the Apollo 1 two additional pounds while pumped full of 100% O2.
Just very weird where people act as if they do not understand the basics while actually operating the real vehicles. I guess it happens more than we would like to think?
p.s . in case anyone hasn't seen it, heres the definitive video. Money shot at about 1:11 and definitely turn on the sound:https://www.youtube.com/watch?v=_BgJEXQkjNQ
Geez, now there are nome conspiracy theorists saying it was attacked by a drone.
http://www.dailystar.co.uk/news/latest-news/542878/spacex-rocket-explosion-elon-musk-falcon-9-attacked
I guess PT Barnum was right.
Snip> I have to say, I don't have a lot of sympathy for Musk and company. They have spent the last 5+ years telling everyone how stupid and backwards everyone else is, particularly the experienced contractors like ULA. I have no great love for ULA, either, but having 50+ years of experience has at least taught them to be very careful. Musk, Rutan, et al seem to operate on hubris. Perhaps they are learning some hard lessons and humility, but no evidence of it so far. <
I've read this in other places also. They really need to suck it up and change their mindset.
NASA and the Aerospace Industry have decades of experience. Mostly success, but they learn from their failures.
The fanbois are even worse. Like this:
http://thefreethoughtproject.com/watch-elon-musk-nasa-school-children-lands-rocket/
NASA and Von Braun considered this during the initial Saturn C1-C5 designs in the late 50's. But they rejected it because it *costs something like 30% of the throw weight* both from the fuel you have to reserve for the landing, and for the landing gear, heat shield, etc, for some hypothetical saving on recurring costs, and the additional risk of flying something multiple times.
Just to be clear, I don't fault them at all for having failures. Search for "early rocket failures" and see how many there were originally. Failures are a necessary element to develop something like this. I do fault them for either not being aware of the vast database of experience, or not knowing about it, all the while claiming to be geniuses for ignoring it.
Brett
Yes, clearly we should use something sensible, like a Lofstrom loop (https://en.wikipedia.org/wiki/Launch_loop). It looks like it needs more power than is available from a small block Chevy engine -- maybe a 427 or two?
The basic NASA problem is that they tended to be just too cautious, and they over-designed, and over-tested everything, until the costs got too much for even our "spend-happy" government to tolerate. So, enter the private sector, saying they could to the same job at far less cost.
Do you know that they did that, or do you just surmise that from the thing's behavior?
'cause, I sometimes advocate seat-of-the-pants loop tuning, but only in cases where the problem is easy to solve and no one is going to get as much as a paper cut if things go haywire. There's no excuse to not do it right (particularly since doing it right followed by 10,000 simulations to verify costs less than 100,000 iterations of doing it wrong).
That's what people very close to the process tell me they did. Hand-tuning the gains is one thing- to some extent, there is always some art to it. It's another to dispense with analytical work entirely and just run a bunch of simulations and hope you hit the right conditions somewhere in there at random.
I know for direct fact that other of the new-space people don't know what they are doing. Carmack was on sci.space. <<something>> explaining why timing didn't matter and Windows RT was plenty good enough for the operating system - for his *gimbal loop* that probably needed to run at 120 hz+. He explained that they had done hundreds of simulations with random timing jitter and thus proved it. OY!
Even the Orion service module, which uses an old OMS pod engine and gimbal control, uses an analog system for the servo loop, partly because no one could build a discrete-sampled system fast enough. It's particularly bad when you take the gigantic OMS engine that used do push around the big old orbiter, and stick it to the back of the relatively tiny Orion CSM.